[ESS] How many ESS/Julia users out there?

[Vitalie Spinu]

 >> Kevin Wright <kw.stat at gmail.com>
 >> on Tue, 13 Aug 2013 11:07:56 -0500 wrote:

 > Looks very interesting.
 > My _impression_ of org-babel is that is more oriented toward publishing,
 > whereas the python notebook looks like it is more focused on an interactive
 > session.  I suggested an R interface like a Mathematica notebook 10 years
 > ago:
 > http://tolstoy.newcastle.edu.au/R/help/02b/4180.htmlI'm glad to see technology may actually make this possible.

Notebook interface can be implemented rather easily on top of our
existent transcript mode. Or even directly on top of ess-mode. I might
even implement it as a byproduct of ess-image mode for viewing process
graphical output directly in emacs, but I am not really interested in
notebook interface per se.

I used mathematica notebooks quite a lot and quite frankly, the
interface sucks. Most of the time you spend on navigating through and
deleting useless output. The way emacs/ESS works, by separating the
source from output, is close to perfect IMO. 

The main issue is indeed with publishing. Often one needs a quick and
dirty presentation of the results for the collaborators, and refactoring
your code into noweb is quite a hustle. For those cases, inline
graphical output in subprocess buffer and then quick export to html or
pdf could be very handy. 

These features are planed and will be in ESS quite soon. It should have
been already there, but it turned out that emacs completely lacks proper
image manipulation API. So I had to write an image-transform library
first. 


  Vitalie